U.S. patent application number 17/558996 was filed with the patent office on 2022-06-23 for methods and systems for large area and low defect monolayer ordering of microspheres and nanospheres.
The applicant listed for this patent is The Board of Trustees of the University of Illinois. Invention is credited to Sartaj Grewal, Lukas Janavicius, Julian Michaels, Dane Sievers.
Application Number | 20220193718 17/558996 |
Document ID | / |
Family ID | 1000006106737 |
Filed Date | 2022-06-23 |
United States Patent
Application |
20220193718 |
Kind Code |
A1 |
Sievers; Dane ; et
al. |
June 23, 2022 |
METHODS AND SYSTEMS FOR LARGE AREA AND LOW DEFECT MONOLAYER
ORDERING OF MICROSPHERES AND NANOSPHERES
Abstract
In an example, a method including dispensing a liquid onto a
first portion of a surface of a substrate and dispensing a solution
comprising colloidal spheres onto a second portion of the surface
of the substrate. The method additionally includes agitating the
colloidal spheres to disperse the colloidal spheres along the first
portion and the second portion of the surface of the substrate and
directing air flow above the colloidal spheres inducing rotation of
the colloidal spheres. In another example, a method includes
positioning a retaining ring on a surface of a liquid above a
substrate below the surface of the liquid and dispensing a solution
comprising colloidal spheres onto the surface of the liquid within
a surface area of the retaining ring. The method further includes
agitating the surface of the liquid and the colloidal spheres to
disperse the colloidal spheres along the surface area of the
retaining ring.
Inventors: |
Sievers; Dane; (Fisher,
IL) ; Grewal; Sartaj; (Champaign, IL) ;
Janavicius; Lukas; (Champaign, IL) ; Michaels;
Julian; (Champaign, IL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
The Board of Trustees of the University of Illinois |
Urbana |
IL |
US |
|
|
Family ID: |
1000006106737 |
Appl. No.: |
17/558996 |
Filed: |
December 22, 2021 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
63129434 |
Dec 22, 2020 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
B05D 1/204 20130101;
C09D 125/06 20130101; B05C 3/05 20130101; B82Y 40/00 20130101; B05D
1/208 20130101; B05C 11/028 20130101; B05C 11/11 20130101; B82Y
30/00 20130101 |
International
Class: |
B05D 1/20 20060101
B05D001/20; B05C 3/05 20060101 B05C003/05; C09D 125/06 20060101
C09D125/06; B05C 11/02 20060101 B05C011/02; B05C 11/11 20060101
B05C011/11 |
Claims
1. A method of assembling a colloidal crystal, the method
comprising: dispensing a liquid onto a first portion of a surface
of a substrate; dispensing a solution comprising colloidal spheres
onto a second portion of the surface of the substrate; agitating
the colloidal spheres so as to disperse the colloidal spheres along
the first portion and the second portion of the surface of the
substrate; and directing air flow above the colloidal spheres
inducing rotation of the colloidal spheres.
2. The method of claim 1, wherein the substrate is a first
substrate, and the method further comprises: transferring the
colloidal spheres from the first substrate to a second
substrate.
3. The method of claim 2, wherein the liquid is a first liquid,
wherein the first substrate and second substrate are mounted on a
platform, and wherein transferring the colloidal spheres from the
first substrate to the second substrate comprises: adding a second
liquid to the platform, such that the colloidal spheres float on
the second liquid above the first substrate; and moving the
colloidal spheres along the surface of the second liquid such that
the colloidal spheres are above the second substrate.
4. The method of claim 3, further comprising a retaining ring
positioned on the surface of the first substrate, and wherein the
first portion and second portion of the surface of the first
substrate are at least partially surrounded by the retaining
ring.
5. The method of claim 4, wherein moving the colloidal spheres
along the surface of the second liquid such that the colloidal
spheres are above the second substrate comprises moving the
retaining ring along the along the surface of the second liquid
such that the retaining ring and the colloidal spheres are above
the second substrate.
6. The method of claim 3, further comprising removing the second
liquid from the platform.
7. The method of claim 1, wherein agitating the spheres comprises
applying acoustic agitation.
8. The method of claim 1, wherein agitating the spheres comprises
mechanical agitation of a surface of the liquid.
9. The method of claim 1, wherein the colloidal spheres are
nanospheres.
10. The method of claim 1, wherein the colloidal spheres are
microspheres.
11. The method of claim 1, wherein the colloidal spheres comprises
polystyrene.
12. The method of claim 1, wherein the solution comprising
colloidal spheres also comprises at least one of: Acetone,
Isopropanol, Methanol, Ethanol, Ethylene Glycol, Propylene Glycol,
Glycerol, and Deionized water.
13. The method of claim 1, wherein the liquid comprises deionized
water.
14. The method of claim 1, wherein the second portion of the
surface of the substrate comprises an edge of the surface of the
substrate.
15. A method of assembling a colloidal crystal, the method
comprising: positioning a retaining ring on a surface of a liquid
above a substrate below the surface of the liquid; dispensing a
solution comprising colloidal spheres onto the surface of the
liquid within a surface area of the retaining ring; agitating the
surface of the liquid and the colloidal spheres to disperse the
colloidal spheres along the surface area of the retaining ring; and
removing at least a portion of the liquid to transfer the retaining
ring and the colloidal spheres onto a surface of the substrate.
16. The method of claim 15, further comprising exposing the
colloidal spheres to a solvent vapor.
17. The method of claim 15, wherein agitating the surface of the
liquid and the colloidal spheres comprises: applying a paddle to
the surface of the liquid by periodically moving the paddle along a
vertical path.
18. The method of claim 15, wherein agitating the surface of the
liquid and the colloidal spheres comprises: periodically moving a
paddle along a horizontal path along the surface of the liquid.
19. The method of claim 15, wherein agitating the surface of the
liquid and the colloidal spheres comprises agitating the surface of
the liquid and the colloidal spheres by way of an acoustic
exciter.
20. The method of claim 15, wherein the retaining ring comprises a
notch extending from a top portion of the ring to a bottom portion
of the ring, and wherein dispensing a solution comprising colloidal
spheres onto the surface of the liquid within a surface area of the
retaining ring comprises dispensing the solution onto the notch of
the retaining ring.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present application is a non-provisional patent
application claiming priority to U.S. Provisional Patent
Application No. 63/129,434, filed Dec. 22, 2020, the contents of
which are hereby incorporated by reference.
FIELD OF THE DISCLOSURE
[0002] The present disclosure relates to large area and low defect
monolayer ordering of microspheres and nanospheres. More
particularly, embodiments of the present disclosure describe
assembly of colloidal crystals.
BACKGROUND
[0003] Nanosphere lithography (NSL) is a low-cost, materials
general, inherently parallel, and high throughput bottom-up
fabrication technique. The characteristic feature of NSL is the
self-assembled colloidal nanosphere mask, which enables a more
cost-effective option for large area patterning compared to serial
fabrication with electron beam lithography. The resultant feature
size is linked to the diameter of a single constituent nanosphere
and can be varied from sub-50 nm to 10 .mu.m. Therefore, NSL offers
a flexible lithography tool for fabricating a number of devices
like surface-enhanced Raman sensors, metasurfaces, photovoltaics,
resistive switching memory, transparent electrodes, light-emitting
diode arrays, chemical sensors, superhydrophobic surfaces and
lasers.
[0004] The colloidal sphere mask can be fabricated through dip
coating, spin coating, convective assembly, Langmuir-Blodgett
assembly or air-water interfacial assembly. Air-water interfacial
assembly holds great promise towards achieving wafer-scale
low-defect single-crystal masks. The assembly method benefits from
the inherent ability of the colloidal spheres to spontaneously
self-assemble into a loose-packed monolayer at the air-water
interface. The self-assembly process is driven by attractive
capillary forces due to the air-water interface depression by the
spheres and resisted by the steric repulsion between the spheres.
Close hexagonal packing is achieved with the addition of
surfactants, which lower the intersphere attraction forces and
allow the spheres to rearrange themselves. The resultant colloidal
masks have demonstrated short-range ordering, and large
centimeter-sized single crystals using external energy sources.
SUMMARY
[0005] Single-layer, periodic repeating geometric patterns of
spheres with micrometer and nanometer dimensions can exhibit very
interesting properties, such as the ability to guide light across a
surface or manipulate electromagnetic radiation. A common example
of this structure, the gemstone opal, is composed of micrometer
dimension spheres of silica (glass) arranged in a repeating
geometric pattern; this pattern interacts with light to create
sparkling patterns of color which are considered by many to be
beautiful. In Langmuir-Blodgett films, organic materials are
assembled as an ordered monolayer on the surface of a liquid. The
Langmuir-Blodgett films are created utilizing a Langmuir-Blodgett
trough.
[0006] The Langmuir-Blodgett trough utilizes on-water monolayer
assembly using the walls of a rectangular or circular container of
liquid to confine the material to be assembled; material is added
to the surface of the water until the water surface is covered
completely, then a movable barrier is utilized to compress the
material to a specified surface pressure. After compression, the
monolayer film is transferred to a solid substrate by immersing the
substrate into the liquid, then withdrawing the substrate at a
normal or near-normal angle to the liquid surface. The
Langmuir-Blodgett method and apparatus rely on the compression of
nanospheres and micrometer scale monolayers to a specified surface
pressure to indicate maximum packing density of the material, and
utilizes only a liquid and the material to be deposited for the
assembly process.
[0007] An example method of the present disclosure includes
dispensing a liquid onto a first portion of a surface of a
substrate. The method further includes dispensing a solution
comprising colloidal spheres onto a second portion of the surface
of the substrate. The method additionally includes agitating the
colloidal spheres to disperse the colloidal spheres along the first
portion and the second portion of the surface of the substrate. And
the method includes directing air flow above the colloidal spheres
inducing rotation of the colloidal spheres.
[0008] In another example, a method includes positioning a
retaining ring on a surface of a liquid above a substrate below the
surface of the liquid and dispensing a solution comprising
colloidal spheres onto the surface of the liquid within a surface
area of the retaining ring. The method further includes agitating
the surface of the liquid and the colloidal spheres to disperse the
colloidal spheres along the surface area of the retaining ring. The
method additionally includes removing the liquid to transfer the
retaining ring and the colloidal spheres onto a surface of the
substrate.
BRIEF DESCRIPTION OF THE FIGURES
[0009] The above, as well as additional, features will be better
understood through the following illustrative and non-limiting
detailed description of example embodiments, with reference to the
appended drawings.
[0010] FIG. 1 illustrates an example embodiment of a purpose-built
assembly step.
[0011] FIGS. 2A-2D illustrate steps of an example purpose-built
assembly process.
[0012] FIGS. 3A-3D illustrate steps of an example self-assembly
process.
[0013] FIGS. 4A-4C illustrate example colloidal solution
regimes.
[0014] FIG. 5 illustrates an example embodiment of the solvent
annealing chamber.
[0015] FIGS. 6A-6D illustrates images of colloidal crystals
throughout the treatment process, according to an example
embodiment.
[0016] FIG. 6E illustrates a graph of defect areas for treated and
untreated colloidal crystals, according to an example
embodiment.
[0017] FIG. 7 illustrates a laser diffraction setup for large-area
twist measurement, according to an example embodiment.
[0018] FIG. 8 illustrates an image of diffraction spots on a
crystal.
[0019] FIGS. 9A-9B maps relative crystal orientation for various
diffraction measurement paths.
[0020] FIGS. 10A-10H illustrate process steps for the On-Wafer
Assembly process flow, according to an example embodiment.
[0021] FIGS. 11A-11B illustrate a liquid float transfer setup and
an on-wafer assembly setup, according to an example embodiment.
[0022] FIG. 12 illustrates an on-water assembly system, according
to an example embodiment.
[0023] FIGS. 13A-13I illustrates retaining ring geometries,
according to example embodiments.
[0024] FIG. 14 illustrates an on-water ordered sphere assembly
setup, according to an example embodiment.
[0025] All the figures are schematic, not necessarily to scale, and
generally only show parts which are necessary to elucidate example
embodiments, wherein other parts may be omitted or merely
suggested.
DETAILED DESCRIPTION
[0026] Example embodiments will now be described more fully
hereinafter with reference to the accompanying drawings. That which
is encompassed by the claims may, however, be embodied in many
different forms and should not be construed as limited to the
embodiments set forth herein; rather, these embodiments are
provided by way of example. Furthermore, like numbers refer to the
same or similar elements or components throughout.
[0027] Nanosphere lithography (NSL) is an inexpensive and powerful
lithography technique that allows large-area parallel fabrication
of sub-diffraction limit features. Widespread application of NSL is
hindered by an incomplete understanding of the NSL mask
self-assembly process. As a result, published mask fabrication
recipes are hard to replicate with variable grain sizes and defect
densities.
[0028] Embodiments of the present disclosure include a dynamic
self-assembly process of approximately 1 .mu.m polystyrene
colloidal spheres. It is observed that addition of propylene glycol
to the colloidal solution, and, use of low-velocity air and
low-frequency acoustic external energy sources provides reliable
fabrication of inch-scale quasi single crystals. The short-range
and long-range defect density characterization identifies sphere
diameter variation and evaporation induced line defects as the
primary causes of defect generation. Solvent treatment of the
colloidal crystal reduces mean defect density by a factor of
4.5.times. with a complete elimination of evaporation induced line
defects.
[0029] However, creating a single layer of ordered spheres across a
large area is not easily achieved. Since the first publication
describing a method for "Natural Ordering" of spheres in 1980, the
largest reported area of self-assembled spheres in a low-defect
hexagonal close-packed geometric arrangement is no larger than 1 cm
in diameter. Larger areas have been reported, but the periodicity
of the arrangement is poor, consisting of small "grains" on the
order of square millimeter (often less) areas which are randomly
oriented to each other, with large defect regions between the
individual grains. In addition, the most common method for single
layer ordering of spheres is the Langmuir-Blodgett technique, which
is dependent on operator skill and specialized dedicated equipment.
However, Langmuir-Blodgett films were initially used for deposition
of organic monolayers onto a solid substrate and only later adapted
for the assembly of micro- and nanospheres. Other techniques, such
as convective assembly, are not suitable for large scale production
due to the very long time scales (days to weeks) required for the
ordering process. Because of these limitations described above, the
application of ordered spheres has not progressed beyond use in
research environments.
[0030] The ability to generate large ordered monolayer films in
short time periods would make practical the ability to improve
existing manufacturing methods or the introduction of new
revolutionary products, such as large area micro-lens arrays for
imaging systems, ultra-high reflectivity meta-materials, photonic
crystals, electromagnetic cloaking structures for stealth
applications, nanowires, addressable field-emission structures, and
large ion thrust propulsion systems, to name a few.
[0031] Embodiments of the present disclosure include methods and
systems that overcome the limitations of producing ordered sphere
monolayers. Embodiments include multiple methods for the production
of highly ordered, low defect, very large area hexagonally
close-packed monolayers of spheres. In embodiments, the methods
have demonstrated the ability to create single uninterrupted grains
exceeding 6 cm in diameter of repeatable quality in less than an
hour. The dimensions achievable have been limited only by the size
of the substrate and equipment used during assembly; as system size
has increased, the resulting grain dimension has scaled linearly.
The defect density of the assembled grain is dependent upon the
method utilized, ranging from `good enough` for many applications
when using the basic method, to defect densities that are
determined only by the quality of the spheres used for
assembly--i.e. individual defects can be attributed traced to a
sphere that is misshaped or mis-sized.
[0032] The characterization of long-range ordering in large single
crystals is lacking. The single crystal formation process under the
presence of internal stresses due to polydisperse spheres and,
external stresses due to the energy sources and the evaporation of
water is not well understood. Additionally, the role of the surface
properties of the colloidal spheres and the air-water interface in
achieving good long-range ordering needs to be evaluated.
[0033] In certain embodiments, the methods utilize a confined water
surface to define the plane where assembly occurs. Additional
mechanisms and chemistries have been developed to control the
surface tension, density, sticking coefficients, external energy
input, internal forces, and air flow of the system. These factors
have been isolated and have been determined to influence the
quality, size, and reproducibility of the assembly process. The
methods developed are low cost and utilize common and nontoxic
materials and chemicals.
[0034] In additional to methodologies for producing large area
assembled spheres, additional embodiments include complete
stand-alone systems to control and automate the assembly process,
remove the requirement for a skilled operator, improve the
repeatability of the resulting assembled grain, and allow for
production in most environments, including large-scale and high
volume manufacturing.
[0035] Now referring to FIG. 1, an example embodiment of a
purpose-built assembly setup 100 to form a colloidal crystal 114.
More particularly, FIG. 1 shows a purpose-built setup 100 providing
control of air velocity and flow pattern, and, acoustic energy
input to optimize the colloidal sphere assembly. In examples, a
purpose-built setup 100 includes an acrylic enclosure 102 including
an inlet fan 104 and an acrylic bench 106. Additionally, in some
examples, the purpose-built setup 100 includes an acoustic actuator
108.
[0036] In an example, a silicon wafer 110, for example, a 100 mm
silicon wafer, is placed in the front-open acrylic enclosure 102
fitted with a variable speed inlet fan 104 and the acoustic
actuator 108. The silicon wafer 110 is completely coated with a
deionized water (DI) layer with the exception of a small edge area
to serve as the colloidal solution dispense site. The surface water
contact angle for the silicon wafer can be increased to about
40.degree. (ranging from 35.degree.-45.degree. using hydrocarbon
deposition for approximately 120 hours in an ISO 1000 cleanroom
environment. This allows higher volumes of DI to be loaded, which
increases the air-water interface curvature and efficiently couples
external energy on to the interface.
[0037] The colloidal solution, hereafter referred to as DIPGIP,
comprises of 5:15:15:10:1 of DI:PG:isopropanol (IPA): 10% wt. 1
.mu.m polystyrene spheres (PSL):methanol. DIPGIP is pipetted in
small volume iterations at the edge dispense site by way of a
pipette 112. The colloidal spheres are slowly picked up by the
water contact front and added to the curved air-water interface of
the DI layer.
[0038] Now referring to FIGS. 2A-2D, steps of an example
purpose-built assembly process. More particularly, FIGS. 2A-2D
illustrate the formation process for an inch-sized quasi single
crystal.
[0039] At the early stages of the solution injection process, as
shown in FIG. 2A, the colloidal spheres form loose chains on the
air-water interface. These loose chains are transported away from
the dispensing site by the Marangoni forces, rotated away from the
left edge by the air currents, and aggregate in the center of the
wafer.
[0040] As the colloidal sphere chain concentration increases, the
loose chains compact to form a polycrystalline grain core, as shown
in FIG. 2B, due to the increased capillary force attraction at the
curved air-water interface and the surface adsorbed PG layer on the
colloidal spheres. The PG coating layer, similar to traditional
surfactants, reduces the local surface tension and improves sphere
mobility. The degree of hexagonal symmetry of the assembling film
is visually identifiable due to the grating-like structure created
by the spheres, with grain orientation defining the observed color.
Grain boundaries cause sharp changes in the diffraction pattern,
thereby providing a visual characterization of the long-range
ordering. It is observed that upon formation of the central
polycrystalline core, dispensing additional colloidal solution
contributes to an increase in the total area of the core, where
arriving spheres either contribute to the main core or produce
smaller grains at the perimeter. Simultaneously, the grains in the
central core coalesce, through grain rotation induced grain
coalescence (GRIGC) resulting from low-frequency acoustic
vibrations supplied to the system. GRIGC involves rotation of the
constituent grains to eliminate grain boundaries, where the
boundary elimination time is inversely proportional to the grain
size. This process of grain boundary reduction is accelerated by
imparting a circular rotation to the polycrystalline core using
directed air currents. FIG. 2C shows the polycrystalline grain core
after the end of the recipe addition process, where a number of
large quasi single crystals are observed.
[0041] The optimization process continues until a single inch-sized
quasi single crystal is formed. At this point, the external energy
sources are switched off. The crystal is extracted by breaking the
DI layer surface at the wafer edge with an absorbent cleanroom wipe
(e.g., Texwipe TechniCloth). This leads to draining of the bulk DI
layer, and, the compaction of the quasi-single crystal due to
Marangoni forces. The resultant inch-sized quasi single crystal is
shown in FIG. 2D. Additionally, a significant number of the spheres
introduced into the bulk volume of the DI layer are removed. These
spheres are known to cause triplet defects in two-dimensional
colloidal crystals. By leveraging the physical boundaries of the
target substrate, the properties of the selected chemistries, and,
the solid-liquid/liquid-liquid/liquid-gas interfaces, this
technique provides the same localization effects that up to this
point in time required the use of a polymer ring. As a result, the
technique reduces process complexity, increases cleanliness of the
resulting structure, and eliminates the defects introduced by
transfer processes.
[0042] The surface properties of the colloidal spheres play a vital
role in the formation of large-area single crystals. It has been
shown that the sphere surface modified with adsorbed alcohols or
surfactants mediates the intersphere attraction forces, thereby
driving the formation of a close-packed colloidal polycrystal.
Single crystal ordering can be subsequently enforced in the
polycrystal with the use of external energy sources. The
requirement for complicated energy sources can, however, be
alleviated by optimizing the colloidal solution. Therefore, in
addition to the commonly used alcohols like IPA and methanol,
DIPGIP also contains PG. The choice for PG is made due to an
equivalent density to polystyrene (1.033 g/cm.sup.3), a reduced
surface tension as compared to water (45.6 dynes/cm at 25.degree.
C.) and a low vapor pressure (20 Pa at 20.degree. C.).
[0043] FIG. 3A-3D show an embodiment of self-assembled colloidal
crystals on air-water interface for DIPGIP solution, as shown in
FIG. 3A, DIPGIP solution with shaking, as shown in FIG. 3B, 4:1
IPA:PSL solution, as shown in FIG. 3C, and 4:1 IPA:PSL solution
with shaking, as shown in FIG. 3D. In embodiments shown in FIGS.
3A-3D, the silicon wafers are 55 mm in diameter.
[0044] FIGS. 3A and 3C show the static self-assembled colloidal
crystals for the DIPGIP solution and a non-PG solution containing
4:1 IPA:PSL respectively. The interfacial assembly is performed
without the presence of external air currents and acoustic
vibrations. It is observed from FIG. 3A that the inclusion of PG
results in millimeter-sized single crystals upon loading. The
surface adsorbed PG layer on the colloidal spheres reduces the
local surface tension, which improves sphere mobility and allows
sphere grains to slide multiple lattice sites in order to prevent
grain boundary formation.
[0045] These polycrystals can be optimized into centimeter-sized
quasi single crystals by a simple shaking motion, as shown in FIG.
3B. The localization effect inherent in the assembly process
preserves the dynamic structure of the large-area crystal even
after the shaking motion is ceased. Meanwhile, the colloidal
crystals from the 4:1 IPA:PSL solution do not optimize into large
single crystals under the same dynamic conditions, as shown in FIG.
3D.
[0046] The surface adsorbed PG layer also enables efficient loading
of the colloidal spheres on to the curved air-water interface. For
example, a colloidal solution droplet upon dispense at the edge of
the silicon wafer. Following the dispense, the droplet expands on
the hydrophilic silicon surface. This leads to the evaporation of
the alcohol in the droplet and retraction of the waterfront in
close proximity to the alcohol vapor. The remaining mixture of DI,
PG, and PSL, agglomerates into smaller droplets due to surface
tension. The mixture droplets are surrounded by fields of monolayer
sphere grains precipitated out during the agglomeration
process.
[0047] Once the water front starts advancing, three different
loading regimes for the colloidal spheres are observed. These
regimes are depicted in FIGS. 4A-4C. Regime 1, as shown in FIG. 4A
is initiated when the advancing water front makes contact with a
mixture droplet. Initially, the droplet influx volume dwarfs the
water front loading rate. As a result, flow vortexes are generated
which leads to an accumulation of sphere multilayers at the
waterfront. The resultant non-optimal loading introduces a large
number of spheres to the bulk volume of water.
[0048] Regime 1 quickly transitions into Regime 2, as shown in FIG.
4B, when an equilibrium is established between the droplet influx
volume and the waterfront loading rate. The turbulent influx flow
in the previous regime transitions into a laminar influx flow and
generates monolayer sphere grains at the pinning point of the
waterfront. The preferential formation of monolayer sphere grains
is attributed to the sphere PG coating as this phenomenon is not
observed for non-PG solutions. The resultant loading is
preferential to the air-water interface and introduces
significantly fewer spheres into the bulk volume of water.
[0049] As the puddle influx slows down further, the loading process
transitions to Regime 3, as shown in FIG. 4C, which is marked by
the formation of monolayer sphere grains away from the waterfront.
These grains agglomerate into larger anisotropic chains upon
departure of their accompanying solvent. The waterfront is observed
to unzip the long chains in a sphere by sphere stripping process
with the majority of the spheres transferring to the air-water
interface. The loading process continues in Regime 3 until the next
puddle is encountered. It is important to note that the loading
process can be optimized to load primarily in Regime 2 with the use
of a syringe pump. The loading regimes for the non-PG solutions can
contain 4:1 IPA:PSL. It is observed that the absence of the
mediating effect of PG leads to frequent formation of sphere
multilayers at the waterfront. The resultant loading introduces a
large number of spheres into the bulk volume of water.
[0050] Solvent annealing treatment of colloidal crystals leads to a
significant reduction in the line defect density. In this
technique, an aerosolized solvent gas is applied to the colloidal
crystal on the air-water interface prior to extraction. The polymer
chains on the spheres absorb the solvent molecules, migrate over to
the contact points with the neighboring spheres and fuse with other
polymer chains to minimize their surface free energy. This results
in the creation of a crack-resistant two-dimensional colloidal
crystal film. The aerosolized solvent vapor is produced by passing
a carrier nitrogen gas through a solvent cement (e.g., Weld-On #4).
The resultant vapor contains trichloroethylene, which readily
dissolves polystyrene. FIG. 5 shows the solvent annealing chamber
that functions as an add-on module for the assembly setup shown in
FIG. 1.
[0051] More particularly, an example embodiment of the solvent
annealing setup 500, as shown in FIG. 5, includes an acrylic
enclosure 502, gas inlets 504, and gas outlets 506. The acrylic
enclosure 502 includes a screen 508 and screen mesh 510 (e.g.,
hexagonal pipe mesh) above the sample 512 (e.g., a large area
colloidal crystal).
[0052] The chamber is designed as a wind tunnel settling chamber
that achieves a low-speed highly uniform laminar flow. The
aerosolized vapor enters the chamber through the gas inlets 504
(e.g., a showerhead) and immediately expands in the volume above
the screen mesh 510. Subsequently, the vapor starts moving down due
to gravity effects. The screen mesh 510 eliminates the longitudinal
flow components and the honeycomb flow straightener laminarizes the
vertical flow components. After application on the large area
colloidal crystal 512, the solvent vapor exhausts out at the base
of the chamber 502 at the gas outlets 506.
[0053] The inner structure of the inch-scale quasi single crystal
is probed using high-vacuum high-energy scanning electron
microscopy (SEM). Large-area SEM scans are captured from randomly
distributed points in the crystal. FIG. 6A shows an SEM image of an
embodiment of the untreated colloidal crystal (Scale: 25 .mu.m) and
FIG. 6B shows an SEM image of a colloidal crystal with 30 minutes
of solvent treatment. FIG. 6C shows a processed image of FIG. 6A.
FIG. 6D shows a processed image of FIG. 6B.
[0054] FIG. 6E shows a graph of defect area for untreated ("X"
markers) and solvent treated ("O" markers) colloidal crystals. Each
data point represents one large area SEM scan. It is observed that
imaging the colloidal crystal leads to SEM edge effects with the
defect sites appearing brighter than the neighboring perfect
crystal. This contrast is attributed to the higher number of
secondary electrons that can escape from the defect sites. As a
result, SEM imaging presents a highly sensitive tool for
qualitative assessment of the crystal defect density. The colloidal
crystal possesses point defects, such as miscoordinated spheres,
monovacancies, divacancies, multilayer triplet defects, and
dendritic line defects.
[0055] Point defects can include miscoordinated spheres and
vacancies are caused by the colloidal sphere size and shape
polydispersity. Triplet defects are generated by submerged spheres
pushing up on the monolayer crystal during the drying process. The
displaced spheres reconstruct into a triplet formation. The
dendritic line defects are caused by the ultra-fast crystal lattice
transformation due to the evaporation of water. Line defects
originate from the point defect sites.
[0056] As noted above, FIG. 6B show a large area SEM scan of a
colloidal crystal with 30 minutes of solvent treatment. The solvent
annealing process largely eliminates dendritic line defects.
However, the point source defects of miscoordinated spheres,
vacancies and triplets are still present. The defect density is
quantified using a MATLAB algorithm derived from the pair
correlation function approach. The large area SEM scans are
processed using circular Hough transform and Delaunay triangulation
to create a triangular mesh connecting the sphere centers. The
distance between neighboring spheres is measured against a range of
105 nm around the mean sphere diameter. If the distance falls
outside the range, then a defect is detected.
[0057] FIG. 6C and FIG. 6D show the defective areas overlaid over
the original images of FIG. 6A and FIG. 6B respectively. FIG. 6E
plots the defect area percentages of large area SEM scans for both
untreated and solvent treated colloidal crystals. It is observed
that the solvent annealing treatment leads to .about.4.5.times.
improvement of the mean defect area percentage. Further, very low
disorder (<1%) colloidal crystals require a strategy for
reduction of the sphere diameter variation.
[0058] Pattern formation is a fundamental feature of dynamic
self-assemblies. The circular symmetric external energy input and a
lack of global boundary condition leads to a circular symmetric
twist in the large quasi-single crystal. FIG. 7 shows the laser
diffraction setup 700 for large-area twist measurement. The laser
diffraction setup 700 includes a laser 702, a screen 704, and a
camera 706. The screen 704 is above the sample 710. In some
examples, the screen 704 includes a hexagonal diffraction pattern
708.
[0059] In examples, the inch-scale quasi single crystal from FIG.
2D, shown in FIG. 7 as element 710, is exposed to a laser 702
(e.g., a HeNe laser) and the resultant diffraction pattern is
recorded on a screen 704. As the laser spot is moved along
cross-section lines over the colloidal crystal (six lines are shown
in FIG. 8), the rotation of the diffraction pattern gives the
change in the relative crystal orientation. The high level of
hexagonal crystal symmetry is demonstrated by the crisp diffraction
spots given in the inset of FIG. 8. Additionally, a qualitative
assessment of the circular symmetric twist can be made by observing
the contours of the reflection patterns in the optical image.
[0060] FIG. 9A maps the relative crystal orientation as the laser
spot moves along the y=0 mm path. The sharp change in the
orientation towards the end of the path indicates a line defect.
This effect can be qualitatively observed with the large line
defect marked in FIG. 8. FIG. 9B maps the relative crystal
orientation for all of the measurement paths. The roughness in the
curves is attributed to image processing errors. The circular
symmetric crystal twist is indicated with the crystal edge having
the maximum relative orientation of 34.degree. with respect to the
crystal center. In order to enforce a single crystal orientation, a
hexagonal boundary confinement can be employed.
[0061] Embodiments of the present disclosure include a reliable
recipe for fabrication of inch-scale quasi single colloidal
crystals on a curved air-water interface. In an embodiment, the
addition of PG to the colloidal solution is shown to effectively
modulate the intersphere attraction forces. This results in the
creation of centimeter-scale quasi single crystals just by shaking
the substrate. Thereafter, regimented external energy inputs in the
form of low-velocity air currents and low-frequency acoustic
vibrations are shown to support grain coalescence. The long-range
crystal defect density is probed using scanning electron
microscopy, laser diffraction and image processing. It is deduced
that the primary causes of crystal defects are the sphere diameter
and shape polydispersity, and, the stresses due to the evaporation
of water. Line defect generation due to the evaporation stresses
can be eliminated using solvent annealing, which results in a
4.5.times. reduction in total defect density and total elimination
of line defects. The presence of circular symmetric twist in the
large quasi single crystal is attributed to the circular symmetry
of the external energy sources and lack of a global boundary
condition. In a broader context, this study lays the pathway for
the realization of wafer-scale low-defect single crystal masks and
wider application of nanosphere lithography.
Examples
A. Method for Production of Low Defect, Conformable, Surface
Independent, Very Large Area Continuous Single-Domain Close-Packed
Self-Assembled Micrometer- and Nanometer-Scale Diameter Spheres
[0062] A method for producing high quality geometrically and
spatially periodic close-packed ordered spheres on arbitrary
surfaces of arbitrary x, y, and z dimensions, surface topology, and
cross-sectional profile is disclosed. An embodiment of the
technique is disclosed below. Additional extensions to the basic
methodology are presented as separate methods, and are utilized for
optimization of resulting ordered films based on user defined
requirements and the intended use and/or purpose of resulting
ordered films.
[0063] The core method for production comprises following materials
and processes:
[0064] Colloidal Spheres to Be Utilized for Ordering: micro- and
nanometer-scale monodisperse spheres which are available in a
variety of material composition, diameters, uniformity of dimension
and shape (coefficient of variance, Cv), and concentration in a
colloidal solution. A common sphere specification (and for which
the majority of development work has been performed, but which the
disclosed method is not limited to) is presented as an example:
[0065] Polystyrene latex spheres [0066] 1 .mu.m diameter [0067]
Cv=1.5% [0068] 10% w/v in deionized water (includes a trivial
amount of surfactant for initial wetting of spheres)
[0069] Note that sphere diameter, material, concentration, and
coefficient of variance influence the chemical constituents of the
diluent and the resulting quality of assembled mono layer. Dilute
Sphere Dispensing Solution: the colloidal sphere solution is
diluted for dispensing using one or more of a combination of
chemicals selected for specific properties such as density, surface
tension, vapor pressure, and viscosity.
[0070] Common diluents include, but are not limited to: [0071]
Solvents: Acetone, Isopropanol, Methanol, Ethanol [0072] Glycols:
Ethylene Glycol, Propylene Glycol, Glycerol [0073] Deionized
Water
[0074] An example dispensing solution utilized for 1 .mu.m
polystyrene consists of the following mixture: [0075] 1 part by
volume commercially sourced 10% w/v 1 .mu.m diameter polystyrene
spheres colloidal suspension in deionized water [0076] 1 part by
volume ACS Reagent Grade methanol [0077] 3 parts by volume ACS
Reagent Grade propylene glycol
[0078] Note that composition of the dispensing solution is
dependent upon sphere diameter, dispensing technique, and
acceptable defect density and defect type. Composition and ratios
of the solution are determined by sphere material, dimension, and
dispense method. All materials utilized are ACS Reagent grade or
more stringent purity standards, including but not limited to
Semiconductor/CMOS Grade and HPLC Grade. High purity materials must
be utilized to prevent contamination to the surface which spheres
will be deposited and to control reproducibility.
[0079] Preparation of Surface onto Which Ordered Spheres Are
Deposited: cleaning and surface modification is dependent upon the
specific ordering technique deployed and is discussed in detail for
the specific methods which follow the main embodiment. Example
surface preparation process utilizing a single crystal silicon
substrate: [0080] Degreasing of substrate utilizing washes with
acetone, isopropanol, deionized water, and/or other solvents to
remove surface organic contaminants [0081] Hydrofluoric acid etch
to remove native surface oxide [0082] Isopropanol rinse followed by
nitrogen blow drying to slightly reduce contact angle of surface;
resulting surface exhibits a contact angle less than original
surface, but greater than a fully hydroxylated surface
[0083] Mechanism for Dispensing Spheres: a device for dispensing
controlled amounts (.mu.L scale) of dispensing solution, such as a
micro-syringe or pipette, is utilized to place the sphere
dispensing solution controllably to begin the assembly process.
[0084] Dispensing onto Liquid Surface: self-assembly of spheres
occurs on the surface of a laterally confined liquid, typically
deionized water. Sphere dispensing solution is applied to the
liquid surface with a syringe or pipette utilizing techniques which
promote floating of spheres on the liquid surface and are dependent
upon the assembly method employed. Based on physical arrangement of
confined liquid surface and composition of dispensing solution,
spheres begin assembling into a single floating mono layer mass.
Additional spheres are added until desired dimension of domain is
achieved; dimension of sphere domain must be less than dimension of
confined liquid surface area to allow unrestrained mobility of
domain while assembling (approximately 2/3 of the area of confined
liquid surface).
[0085] Controlled Input of Energy to Promote Assembly: energy is
supplied to the system in order to overcome potential energy
barriers which prevent the propagation of ordering processes.
Controllable sources of energy input identified to influence the
assembly process include: [0086] Acoustic excitation (amplitude,
frequency, signal shape) [0087] Mechanical agitation of liquid
surface (amplitude, frequency, direction) [0088] Electrostatic
potentials (charged surfaces, dipoles) [0089] Fluid boundary layer
conditions (air flow, drag forces) [0090] Surface energy [0091]
Convective motion [0092] Centrifugal/Centripetal forces [0093] Bulk
material/surface force vectors (direction of force exerted on
spheres from fluid boundaries)
[0094] Energy is supplied to the floating spheres until the desired
degree of ordering is observed, upon which time the input of energy
is terminated. Typical assembly times vary from 15 minutes to
approximately 1 hour and are dependent upon surface area of
assembly and specific method of assembly employed.
[0095] Removal of Spheres Suspended in Bulk of Liquid (Optional):
During the dispensing process, some fraction of spheres are
introduced into the bulk of the liquid used for assembly. These
spheres do not participate in the ordering process as they are
below the surface where ordering occurs. After assembly and during
transfer of assembled spheres to the desired substrate, suspended
spheres are entrapped between the substrate and ordered monolayer,
introducing three dimensional defects (termed `triplets`) in the
finished assembly. To reduce these defects, suspended spheres may
be removed by exchanging the liquid used during assembly with clean
liquid prior to final deposition onto the desired substrate.
[0096] Solvent Anneal (Optional): After assembly and termination of
energy input, an apparatus for injecting a compatible solvent vapor
is utilized to solvent "weld" (fuse) individual spheres into a
contiguous layer. The degree of solvent welding is controlled to
allow spheres to fuse only at tangential points of contact between
the spheres, and to maintain original sphere dimensions.
[0097] Deposition of Ordered Sphere Mono layer after Assembly:
Spheres are transferred to desired substrate by controlled removal
of the liquid used for sphere ordering. Removal is accomplished via
several methods, including but not limited to: evaporation,
draining using a valve incorporated into the system, and absorption
by a suitable material. The ordering process is independent of the
deposition process, allowing for deposition of the ordered
monolayer of spheres onto most surfaces regardless of dimensions,
surface topology, surface cross-sectional shape, and surface
contact angle. Optionally, a compaction process may be employed in
which a "squeegee" or suitable straight edge is placed into contact
with the substrate surface and pushed laterally against the
boundary of ordered spheres. This process can decrease spacing
between the ordered spheres and improve the packing density.
B. Method: On-Wafer Assembly
[0098] A low-cost method requiring common laboratory equipment for
large area assembly of micrometer- and nanometer-scale spheres is
disclosed. This method for sphere assembly exhibits a wide process
window which results in large area single domain (>6 cm diameter
demonstrated) assemblies in time scales of less than one hour, and
can be easily reproduced in most laboratory environments with a
minimal investment in equipment and training, but the resulting
ordered monolayer exhibits higher defect density than other methods
disclosed. This method provides a universally accessible technique
that produces ordered assemblies directly on a surface of interest
at a scale not currently achievable and with defect densities
acceptable for use in most laboratory and research environments.
The dominant defects include stacking ("triplets"), point (missing
spheres), and line (slight displacement of adjacent ordered areas
resulting in a discontinuity in the form of a line of spheres not
touching tangentially at three points). Of these defects, on-wafer
assembly results in higher density of "triplet" stacking defects as
compared to other methods of assembly.
[0099] The on-wafer assembly method utilizes the core process steps
described in Section A. above, but uses further steps as well.
[0100] FIGS. 10A-10H demonstrate the process steps that are
disclosed below for the On-Wafer Assembly process flow.
[0101] 1. Substrate: any clean substrate may be used for the
ordering process; of importance is the contact angle of water on
the surface. Optimal contact angles are between that of a
hydrophobic surface (CA>90.degree.) and a hydrophilic surface
(CA<90.degree.). For reference, a 2'' diameter substrate should
require between 1 mL-3 mL of deionized water to cover the wafer
from edge to edge; a contact angle which allows .about.1 mL of
water to completely, or substantially completely, cover the
substrate is optimal.
[0102] 2. Substrate Surface Preparation: substrate surface should
be clean and particulate free. If the contact angle of the surface
exceeds limits defined in Step 1, additional surface treatments,
such as piranha etch, base piranha etch, partial hydroxylation with
alcohols, oxygen/ozone plasma, UV light irradiation, etc. can be
utilized to achieve required contact angle.
[0103] 3. Sample Platform: The substrate is placed onto a level
perforated platform that is elevated above a solid surface to allow
air to flow above and below the sample; perforations in the
platform introduce turbulence in the air as a mechanism for
introducing energy to the system.
[0104] 4. Dispense Liquid onto Substrate for Surface Assembly, as
shown in FIG. 10A: FIG. 10A illustrates a clean substrate 1002
after dispensing deionized water, leaving a small area of substrate
exposed for sphere solution dispensing. The liquid utilized must
fulfil the density, viscosity, vapor pressure, and surface tension
requirements for the spheres to be assembled. For example,
deionized water is ideal for polystyrene as it has a higher density
than the spheres with sufficient surface tension which promotes
floating of spheres on the surface; the dispense solution has a
different surface tension than water, which is also a requirement
for proper assembly. Sufficient liquid must be dispensed onto the
substrate to cover .about.90% of the substrate surface.
[0105] 5. Dispense Dilute Colloidal Sphere Solution onto Substrate
(FIG. 10B): using a pipette or syringe, the sphere solution is
dispensed onto the area of the substrate not covered with liquid.
Typical quantities dispensed for a 2-inch diameter substrate is on
the order of 3-10 .mu.L, but can vary depending upon substrate
size, dispense solution composition, and contact angle of the
surface. Due to the difference in surface tension between the
liquid used for floating the spheres and the dispense solution, the
liquid on the surface is "pushed" away from the dispense solution
(Marangoni Effect) (as shown in FIG. 10C). This receding of the
liquid allows the dispensed spheres to spread out from the point of
dispense; the rate of spread and drying rate are adjusted by the
composition of the dispense solution, and are optimized to allow
spheres to be dispersed as a thin layer in close proximity to the
liquid on the surface, and the rate of evaporation determines the
time required for the liquid to begin to reflow back to the
original fill area. As the liquid returns, spheres are lifted onto
the surface of the liquid and float (as shown in FIG. 10D). This
process of dispense and float is repeated until approximately 2/3
of the liquid is covered with floating spheres (as shown in FIG.
10E).
[0106] 6. Optimization (FIG. 10F): After loading the liquid surface
with floating spheres, energy is supplied to the system to overcome
local potential energy barriers and allow spheres to assemble into
a periodic close-packed geometric arrangement. At a minimum, two
forms of energy input are required for successful assembly: [0107]
Mechanical Vibrational Energy: the system is provided with low
frequency (28-44 Hz), low amplitude (.about.1.5.times.10.sup.-4
m.sup.2/s.sup.3) vibrational energy in x, y, and z directions.
Transducers for inputting vibrational energy include piezoelectric
devices, acoustic exciters, and mechanical vibrators. [0108]
Centrifugal/Centripetal Energy: air flow over the sample is
directed over the sample with a gradient in face velocity
(.about.20 fpm at surface of sample to .about.85 fpm several inches
above the sample) and in a manner which induces rotation of the
floating sphere mass. Mechanisms for introducing this energy
include: fume hoods, fans, mixing tables; rotation can be induced
through mechanical barriers to direct airflow, fan vectoring, and
nutation of mixing tables.
[0109] 7. Depositing Assembled Spheres onto Substrate (FIGS.
10G-10H): After spheres have assembled to required dimension and
quality, the assembled spheres on the liquid covering the substrate
can be deposited onto the substrate via multiple methods: [0110]
Evaporation: the liquid is allowed to evaporate at room temperature
or can be accelerated by elevating the temperature (below the Tg of
the sphere material) on a hotplate or in an oven. [0111] Wicking:
utilizing a clean non-linting absorbent material (cleanroom wipe or
equivalent, swab) to wick away excess liquid. This technique has
the advantage of reducing suspended spheres which create "triplet"
defects, but also requires careful control in order to minimize
damage to the ordered domain due to potential physical contact with
spheres.
[0112] FIG. 10G illustrates an approximately 75 mm domain. No grain
boundaries are present in the large central domain, although
"twisting" of the crystal is visible. FIG. 10H illustrates
approximately 50 mm.times.80 mm rectangle, single domain with a
lower degree of "twisting" relative to FIG. 10G. This indicates
fewer defects in crystal.
C. Method: On-Wafer Assembly with Liquid Float Transfer
[0113] FIG. 11A shows an on-wafer assembly 1100A. As described in
more detail below, the on-wafer assembly 1100A includes a first
substrate 1102 and a second substrate 1104. Spheres 1106 on the
first substrate 1102 are surrounded by a ring 1108. The ring 1108
can include materials such as Kapton, silicone, etc.
[0114] FIG. 11B shows the float transfer method 1100B. As shown in
FIGS. 11A-11B, the Liquid Float Transfer method is substantially
similar to On-Wafer Assembly, differing only in the last step of
depositing the ordered domain onto the substrate surface. With
On-Wafer Assembly, the liquid surface on which the spheres are
floated and ordered is confined by the edges of the substrate onto
which the ordered assembly will deposit.
[0115] The Liquid Float Transfer Method offers two mechanisms for
improving the On-Wafer Assembly process: [0116] Spheres can be
deposited onto any substrate regardless of the surface contact
angle, dimensions, surface topology, or overall geometry. [0117]
Suspended spheres are reduced in density via dilution of the liquid
used for assembly after ordering is complete.
[0118] Spheres are deposited and optimized as in the On-Wafer
Assembly Method up to and including step 6: Optimization (FIG.
11A). The method is modified as follows:
[0119] 7. A retaining ring with an inside diameter smaller than the
substrate diameter is applied to the substrate prior to dispensing
water and sphere solution. The retaining ring is selected so as to
float on the surface of water, and is utilized for stabilization of
assembled spheres during the transfer process.
[0120] 8. Dilution and Lift-Off: Spheres are ordered upon or
transferred to a solid surface. Liquid of the same type as used for
the ordering process is slowly and carefully added to the perimeter
of the wafer until a sufficient amount has been added to allow the
liquid to extend beyond the bounds of the substrate used for
ordering. Additional liquid can continue to be added in order to
dilute the density of spheres suspended in the liquid. Excess
liquid can be removed as additional liquid is introduced to further
dilute the suspended spheres.
[0121] 9. Transfer Assembled Spheres (optional): a separate clean
substrate is placed onto the transfer platform and additional
liquid is added so that the second substrate is covered with liquid
and extends beyond the bounds of the substrate. Liquid is continued
to be added until a "bridge" forms, connecting the liquid pools
from both substrates. The ordered assembly can be mechanically
manipulated and moved across the liquid bridge to the clean
substrate using the retaining ring.
[0122] 10. Depositing Assembled Spheres onto Second Substrate:
using the same technique described in Step 7 in the previous
method, the assembled spheres can be deposited onto the second
substrate. The retaining ring is removed after transfer.
D. Method: On-Water Assembly
[0123] On-Water assembly is an assembly technique and is an
extension of the on-wafer method for sphere assembly optimized to
allow for depositing ordered sphere assembly onto virtually any
surface regardless of contact angle. In addition, defect density is
reduced, virtually eliminating stacking defects (triplets).
Optimized ordered grain defect density is limited by the Cv of
spheres utilized; defects observed in assembled monolayer can each
be attributed to an individual over-sized, under-sized, or
misshaped sphere. However, the domain can exhibit "twist"; although
the assembled spheres are arranged as a single contiguous domain,
the typical round arrangement of the ordered domain can result in
line defects, which over the large dimensions of the ordered domain
appear as "waviness".
[0124] FIG. 12 illustrates an On-Water Assembly system and FIG. 13
illustrates retaining ring geometries used for dispensing
spheres.
[0125] Referring to FIG. 12, an On-Water Assembly system 1200
includes a substrate 1202 mounted on a stand 1204 within a
container 1218. The substrate 1202 and stand 1204 are submerged in
a liquid, such as water, between vertical columns 1214. The liquid
can be continually injected into the system 1200 by way of an inlet
1208 and removed from the system by way of an outlet 1210. The
liquid can be injected into the substrate container by way
injection pumps 1216. As described in more detail below, the
spheres 1220 can be dispensed onto the surface of the water inside
a retaining ring 1212.
[0126] In some examples, the On-Water Assembly system 1200 includes
an acoustic exciter 1206 for agitating the water and spheres. The
On-Water Assembly system 1200 can be used for On-Water Assembly in
the method described below:
[0127] On-Water Assembly comprises the following apparatus and
materials:
[0128] 1. Container to hold water, with provisions for:
[0129] a. A continuous flow of clean deionized water from the
bottom controlled by a valve. Port dimensions and angle can be
varied to induce various water flow patterns and are tuned to
optimize assembly.
[0130] b. An overflow and catch basin for collecting and draining
water as it is flushed through the system.
[0131] c. A mechanism for draining water from the assembly tank to
allow deposition of assembled spheres onto a substrate submerged
below the surface of water where assembly occurs (FIG. 8d).
[0132] d. A holder with provisions for securing the substrate where
assembled spheres will be deposited and positioning the substrate
below the surface where assembly occurs (1204).
[0133] e. A mechanism for maintaining the position of a ring used
to retain spheres on the surface of the water during assembly.
[0134] 2. Controlled energy input via a combination of one or more
of the following mechanisms:
[0135] a. A paddle (such as paddle 1222 consisting of a vertical
rod attached to a horizontal plate). The paddle is positioned on
the surface of the water, and is moved vertically in a periodic
manner to agitate the surface of the water where spheres are
assembled. The vertical displacement is variable and tuned to the
sphere system, with typical values of 1 mm to 5 mm vertical travel
at a rate of 0.25 Hz to 5 Hz.
[0136] b. A paddle consisting of a horizontal bar attached to a
vertical paddle. The paddle is positioned on the surface of the
water, and is moved horizontally in a periodic manner to agitate
the surface of the water vertically where spheres are assembled.
The vertical displacement is variable and tuned to the sphere
system, with typical values of 1 mm to 5 mm horizontal travel at a
rate of 0.25 Hz to 5 Hz.
[0137] c. One or more acoustic exciters 1206 mounted at various
positions on the container. The exciter is driven using various
waveforms, including but not limited to sinusoidal, pulsed, "heart
beat", saw-tooth, square-wave, and white noise. Amplitude and
frequency of signal is tuned to the sphere system; typical
frequencies range from 1 Hz to 10 s of kHz.
[0138] d. Manual physical motion induced by periodically tapping
the sides of the container.
[0139] 3. Sphere Retaining Ring (as shown in FIGS. 13A-13I). Sphere
assembly requires a confined area for optimal assembly and to allow
for positioning the assembled spheres during the deposition
step.
[0140] a. The retaining ring material and geometry is selected so
that it floats on the surface of water. Materials utilized include,
but are not limited to, polyimide film (for example, Kapton.RTM.),
acrylic (for example, PMMA), acetal (for example, Delrin.RTM.),
PTFE (for example, Teflon.TM.), and silicone rubber (for example,
PDMS).
[0141] b. Retaining ring inner geometries include circular (1300A,
as shown in FIG. 13A), triangular (1300B, as shown in FIG. 13B),
rectangular (1300C, as shown in FIG. 13C), hexagonal (1300D, as
shown in FIG. 13D), and octagonal (1300E, as shown in FIG.
13E).
[0142] c. The ring can float either from the bottom (1300F, as
shown in FIG. 13F) or top surface (1300G, as shown in FIG. 13G) of
the material, and is determined by the density, geometry, contact
angle of water with the material, and placement of the retaining
ring onto the surface of the water. Floating position and contact
angle is determined by the sphere system to be assembled, as both
properties influence the force vectors of the water acting on the
spheres.
[0143] d. The retaining ring can be modified to facilitate
dispensing the sphere solution. Sphere solution can be dispensed
onto the surface of the ring; small cuts through thin retaining
ring material (FIG. 13H) at various angles influence the direction
and speed of the spheres as they float onto the water surface, and
can be used to optimize the initial ordering process. For thicker
materials, the retaining ring can be notched in a manner that
creates a "ramp", where the bottom of the notch is in close
proximity to the water surface (FIG. 13I). The dimensions and angle
of the of the ramp can be adjusted to modify the apparent contact
angle of the sphere solution with the water and can be utilized to
influence the ratio of spheres which float on the surface of the
liquid to spheres which go into suspension below the surface of the
water.
[0144] 4. Sphere Dispensing Solution. The sphere dispensing
solution utilizes the same chemistries as used for the On-Wafer
Assembly method, namely alcohols, glycols, and spheres suspended in
deionized water. The On-Wafer Assembly sphere dispense solutions
have proven to provide high quality sphere assembly. However, a
higher concentration of glycols results in higher quality sphere
assemblies. A specific formulation which demonstrates superior
results is provided as an example: [3:1:1 propylene glycol:
methanol: 10% w/v sphere solution].
[0145] 5. Sphere Dispensing Method and Mechanism. The sphere
solution is dispensed utilizing various methods as described
below.
[0146] a. Dispensing sphere solution directly onto the water using
a pipette. This is a simple technique, but results in a high
concentration of spheres going into suspension below the surface of
the water and requires larger volumes of sphere solution to achieve
the same grain dimensions relative to other techniques.
[0147] b. Dispensing sphere solution directly onto the retaining
ring. This technique requires the use of a very thin retaining
ring, otherwise the spheres do not make contact with the surface of
the water used for assembly. Results can be improved by cutting a
partial line into the retaining ring which extends and is
terminated by the interior edge of the ring. This promotes
dispensing of the sphere solution through capillary action, much
like the nib of a fountain pen. In addition, the angle of the line
cut into the ring influences the initial assembly and can be
optimized for specific spheres.
[0148] c. Dispensing sphere solution onto notched retaining ring.
For thicker retaining rings, a notch can be fabricated which
downward toward the interior of the ring. This allows sphere
solution to come into direct contact with the water surface.
Contact angle of water with retaining ring material is critical and
influences the ratio of spheres which float on the surface of water
to spheres suspended below the surface of the water.
[0149] d. Dispensing sphere solution directly onto water surface
using a micro-syringe and syringe pump (FIG. 14). The syringe is
mounted rigidly with provisions for adjusting x, y, and z position
for optimizing dispensing efficacy and sphere assembly. Needle gage
and shape (blunt, angled, etc.) are selected for optimal dispense
and assembly based on sphere system.
[0150] The On-Water assembly method utilizes the core process steps
described in Section A. above. The complete process sequence is
disclosed below. A functional system with resulting on-water
ordered sphere assembly 1400 is shown in FIG. 14.
[0151] An example on-water ordered sphere assembly 1400 includes a
syringe 1402 above a support base 1404 and a trough 1406. The
syringe is controlled by an automated system pump
positioning/dispensing mechanism 1408. The automated system pump
positioning/dispensing mechanism is attached to a syringe plunger
motor for controlled rate of sphere solution dispensing 1410 and
syringe positioning motors 1412 for placement of the syringe tip on
the surface of the liquid. The on-water ordered sphere assembly
1400 may additionally include a syringe positioning base 1414. In
some examples, an on-water ordered sphere assembly 1400 may
additionally include an annealing solvent injector 1416. The
on-water ordered sphere assembly 1400 further includes a sample
pedestal 1418, as well as variable pitch water injection ports
1420, a water inlet 1422, and a water outlet 1424.
[0152] 1. Initial Conditions. On-Water Assembly system drained and
dry. Mount clean substrate onto substrate holder securely to
prevent movement during subsequent procedures.
[0153] 2. Fill Container with Deionized Water. Water is introduced
at a controlled rate to fill the Assembly container. Water flow is
adjusted to maintain a constant level with minimal disturbance of
the surface.
[0154] 3. Lower syringe tip until it is just touching the water
surface without breaking the surface.
[0155] 4. Using a syringe pump, sphere solution is dispensed onto
the surface of the water at a rate of 5 .mu.L/min-10 .mu.L/min
until desired area is covered.
[0156] 5. After dispensing spheres, external energy sources are
turned on and water flow is adjusted for optimal assembly
conditions.
[0157] 6. Allow spheres to self-assemble; typical ordering times
are between 20 min-30 min. After optimal assembly is achieved,
external energy sources and water flow are turned off.
[0158] 7. Optional: Solvent Anneal Spheres utilizing vaporized
solvent suitable for softening the sphere material. For polystyrene
spheres, typical suitable solvents include, but are not limited to,
toluene, methanol, methanol, methylene chloride. Solvent is placed
into a clean bubbler using nitrogen as the carrier gas. The
vaporized solvent is transported by clean tubing which is attached
to the lid of the system (for example, lid 1224). The solvent vapor
is applied for an appropriate time (typically ranging 30 seconds to
3 minutes) to soften the spheres and allow them to fuse together at
points of contact. The solvent vapor is then removed from the
system.
[0159] 8. Deposit spheres onto substrate. The water drain valve is
opened slowly to allow water level to lower in the container. The
assembled spheres are positioned manually over the substrate as the
water level approaches the level of the substrate surface. The
assembled spheres deposit on the substrate surface; the remainder
of the water in the container is allowed to drain from the
system.
[0160] 9. Drying and removal. The deposited sphere assembly is
allowed to dry and the retaining ring is removed.
[0161] Compared to the Langmuir-Blodgett method, the method and
apparatus disclosed herein provides additional degrees of control
to facilitate the ordering process, including surface energy
modification through the addition of more complex chemistries to
manipulate the contact angle of the liquid on a surface, the
surface tension of the liquid, lubricity and residence time of the
liquid dispense solution on the surface, and controlled segregation
and mixing of different liquids to facilitate dispensing the
solution onto the liquid surface and provide mechanisms which
promote/enhance the ordering process. In addition, the methods and
apparatus disclosed also allow for controlled input of energy into
the system to promote/enhance the ordering process, such as
acoustic excitation, gas-liquid boundary layer drag forces,
periodic perturbations of the liquid surface, convective and fluid
motion of the liquid by selection of mixtures of liquids with
appropriate evaporation rates, dynamic liquid flow patterns through
the use of variable angle water inlet nozzles,
centrifugal/centripetal forces from rotation of the liquid,
potential energy barrier lowering between particles through the use
of localized surface tension gradients induced by selection of
specific chemistries, force-vector control of liquid through the
use of containment rings and the manner in which the containment
rings are floated on the surface of the liquid, and controlled gas
flow velocity gradients above the surface of the liquid. The
results of the Langmuir-Blodgett method and apparatus are inferior
in both dimensional scale and defect density compared to the
results utilizing the methods and apparatus disclosed
[0162] To date, single-domain ordered sphere assembly has been
limited to 10 mm diameter or smaller grains, with high defect
densities. The methods and apparatuses disclosed herein include
embodiments that demonstrate single-domain ordered sphere
assemblies >80 mm diameter, the elimination of defects induced
by the assembly method and apparatus, and theoretical minimum
defect density achievable based on the quality of the spheres used
for assembly--defects are induced only by misshaped spheres or the
substrate geometry. The ordered single-domain sphere assemblies
appear to scale linearly with substrate dimensions with the maximum
grain diameter demonstrated to be approximately 2/3 of the diameter
of the substrate irrespective of actual substrate dimension.
[0163] Unless otherwise explained, all technical and scientific
terms used herein have the same meaning as commonly understood by
one of ordinary skill in the art to which a disclosed disclosure
belongs. The singular terms "a," "an," and "the" include plural
referents unless context clearly indicates otherwise. Similarly,
the word "or" is intended to include "and" unless the context
clearly indicates otherwise. "Comprising" means "including"; hence,
"comprising A or B" means "including A" or "including B" or
"including A and B."
[0164] All references cited herein are hereby incorporated by
reference to the extent not inconsistent with the disclosure
herewith. Although the description herein contains many
specificities, these should not be construed as limiting the scope
of the disclosure but as merely providing illustrations of some of
the presently preferred embodiments of the disclosure.
[0165] The terms and expressions which have been employed herein
are used as terms of description and not of limitation, and there
is no intention in the use of such terms and expressions of
excluding any equivalents of the features shown and described or
portions thereof, but it is recognized that various modifications
are possible within the scope of the disclosure. Thus, it should be
understood that although the present disclosure has been
specifically disclosed by preferred embodiments, exemplary
embodiments and optional features, modification and variation of
the concepts herein disclosed may be resorted to by those skilled
in the art, and that such modifications and variations are
considered to be within the scope of this disclosure as defined by
the appended aspects. The specific embodiments provided herein are
examples of useful embodiments of the present disclosure and it
will be apparent to one skilled in the art that the present
disclosure may be carried out using a large number of variations of
the devices, device components, methods steps set forth in the
present description. As will be obvious to one of skill in the art,
methods and devices useful for the present methods can include a
large number of optional composition and processing elements and
steps.
[0166] Whenever a range is given in the specification, for example,
a temperature range, a time range, or a composition or
concentration range, all intermediate ranges and subranges, as well
as all individual values included in the ranges given are intended
to be included in the disclosure. It will be understood that any
subranges or individual values in a range or subrange that are
included in the description herein can be excluded from the aspects
herein.
Additional Examples
[0167] Silicon Wafer Surface Treatment: 100 mm silicon <100>
wafers (500 .mu.m thick) are RCA cleaned and stored in an ISO 1000
cleanroom environment for 120 hours. The resultant hydrocarbon
deposition on the silicon oxide surface leads to an increase in
water contact angle to .about.40.degree.. The silicon wafers are
degreased before use.
[0168] Colloidal Solution: The colloidal solution is prepared by
mixing 5:15:15:10:1 parts of deionized water: propylene glycol:
isopropanol: 10% wt. 1 .mu.m polystyrene spheres (Alfa Aesar Inc.):
methanol respectively. The solution is ultrasonicated for 10
minutes before use.
[0169] Colloidal Mask Fabrication: The surface-treated silicon
wafer is placed inside the purpose-built colloidal assembly setup
shown in FIG. 1a. The front-open setup enclosure is cut from 3/16''
thick acrylic sheets (Mcmaster Carr Inc.) using a laser cutter and
glued together. The setup enclosure also houses an inlet fan
(Thermaltake CPU fan) located on the back wall and an acoustic
exciter (Dayton Audio 40 W 4.OMEGA. subwoofer) located next to the
wafer bench. The silicon wafer is coated with 10-11 ml of deionized
water leaving a small edge pocket to serve as the solution dispense
site. A colloidal solution volume of 105 .mu.l is slowly dispensed
using a pipette at the edge dispense site over 60 minutes. After
the dispense of the first solution droplet, the external energy
sources are switched on. The air velocity at the face of the fan is
set at 2.5-2.6 m/s. The acoustic exciter is driven using a function
generator with 44 Hz, 1 Vpp and 5 ms square pulses. Following the
solution addition process, the colloidal crystal is allowed to
optimize for 210 minutes. Finally, the colloidal crystal is
extracted by breaking the water surface tension using a
6''.times.6'' Texwipe. The colloidal crystal is allowed to dry in
air for a few hours before moving the substrate.
[0170] Solvent Annealing: Optimized colloidal crystals are solvent
treated for 30 minutes using the home-built solvent annealing box.
Tricholoroethylene is inlet into the solvent annealing box by
bubbling nitrogen gas at 2 scfh through a glass bubbler containing
Weld-On 4 acrylic glue. The solvent annealing box screen mesh
porosity is set at 0.64 and the hexagonal pipe length is set at 7
times the pipe diameter. These parameters ensure a slow
gravity-assisted laminar flow for the solvent vapor. The crystal is
extracted using the same procedure listed above.
[0171] Microscopy: The interfacial assembly is performed on a
silicon wafer placed on the microscope stage of an Olympus BH2
optical microscope. In the colloidal solution loading process, as
detailed in FIG. 3, scans are acquired at 250.times.
magnification.
[0172] Defect Measurement and Analysis: Large area (6144.times.4415
pixel density) scanning electron microscopy scans are obtained at
random points in the quasi single crystals using the FEI Quanta FEG
450 ESEM. The scan parameters are high-vac mode (10.sup.-6 torr),
10 kV accelerating voltage, Everhart-Thornley secondary electron
detector, .about.5 mm working distance, 800.times. magnification,
20 .mu.s dwell time and 3.0 spot size. Samples are not coated with
a metal layer for SEM imaging.
[0173] Image processing is performed on the large area SEM scans to
determine the lithographic defective area for nanosphere
lithography processes using the quasi single crystals. The SEM
scans are processed with a circular Hough transform and Delaunay
triangulation to create a triangular mesh connecting the sphere
centers. The center-to-center links in the triangular mesh are
measured against a range of 105 nm around the mean sphere diameter.
If this condition is not met, then the triangle with the defective
link is marked defective as a whole. Subsequently, the total
defective area of the large area SEM is computed by the sum of the
areas of the defective triangles.
[0174] Twist Measurement and Analysis: The large area twist
measurement setup is shown in FIG. 7. The silicon wafer with the
inch-scale quasi single crystal is rested on a manual microscope
XY-stage. A Melles Griot 5 mW HeNe laser is mounted vertically,
approximately 40 cm above the microscope stage, pointing downwards.
A white paper screen with a circular hole, 4 mm in diameter, is
fixed 5 mm above the microscope stage in such a way that the HeNe
laser beam passes exactly through the 4 mm hole in the screen. The
silicon wafer, sitting on the microscope stage, is the only moving
object in the described experimental setup. A single lens and a
Point Grey CCD camera are mounted above the screen and capture
.about.20 mm.times.20 mm of the central screen area onto the
camera.
[0175] A transformation function is applied to the captured spot
patterns to account for the camera tilt. Subsequently, a ring
enclosing the hexagonal spots is spliced along the .theta. polar
coordinate. The six-spot pattern leads to six peaks in the
intensity vs .theta. coordinate graph. Finally, the twist in the
crystal is determined by the change in the .theta. coordinate of a
chosen single spot as the laser moves through the colloidal
crystal.
[0176] Although the present disclosure has been described with
reference to certain embodiments thereof, other embodiments are
possible without departing from the present disclosure. The spirit
and scope of the appended aspects should not be limited, therefore,
to the description of the preferred embodiments contained herein.
All embodiments that come within the meaning of the aspects, either
literally or by equivalence, are intended to be embraced therein.
Furthermore, the advantages described above are not necessarily the
only advantages of the disclosure, and it is not necessarily
expected that all of the described advantages will be achieved with
every embodiment of the disclosure.
[0177] While some embodiments have been illustrated and described
in detail in the appended drawings and the foregoing description,
such illustration and description are to be considered illustrative
and not restrictive. Other variations to the disclosed embodiments
can be understood and effected in practicing the claims, from a
study of the drawings, the disclosure, and the appended claims. The
mere fact that certain measures or features are recited in mutually
different dependent claims does not indicate that a combination of
these measures or features cannot be used. Any reference signs in
the claims should not be construed as limiting the scope.
* * * * *